This study aims to explore the construct structure of the Turkish version of the "Myself-As-A-Learner Scale" ("MALS"), which was designed to measure academic self-concept. A translated version of MALS was administered to 483 participants (mean age = 14.31, SD = 2.08) in six different cities across Turkey. An exploratory factor analysis revealed a five-factor structure, explaining 50.99% of the total variation in MALS scores. A confirmatory factor analysis indicated acceptable levels of goodness of fit for the five-factor model. The instrument appeared to distinguish contrasting groups of students according to their English, Turkish, and Maths scores, with more successful students reporting higher MALS scores. The MALS scores also attained significant moderate correlations with achievement in those courses. Further, the instrument yielded a high internal consistency (a =0.82). The study concludes that MALS can be an effective tool in endeavours to explore academic self-concept in relation to various related constructs.